Shared manufacturing-based distributed flexible job shop scheduling with supply-demand matching

Abstract

Shared Manufacturing (sharedMfg) is a peer-to-peer (P2P) paradigm for sharing manufacturing resources, derived from the sharing economy. In sharedMfg, manufacturing service resources are integrated into a large set with defined service types. Based on its resource organization structure, we build a model for the shared manufacturing-based distributed flexible job shop scheduling problem (SM-DFJSP) with supply-demand matching. The goal is to minimize both the total cost and the makespan. The SM-DFJSP model enables the scheduling of jobs requiring different manufacturing services across distributed, heterogeneous, and flexible service resource units (SRUs) with diverse manufacturing functions. To solve the SM-DFJSP, we propose a hybrid estimation of distribution algorithm and Tabu search (EDA-TS), including EDA and TS components. Additionally, a multi-populations strategy and non-dominated solutions memory mechanism are designed to improve the exploration ability of the algorithm. Within the EDA component, three probability distribution models and some dispatching rules are designed to generate a new population. In the TS component, three neighborhood search structures are built that adopt hybrid short and long memory tabu strategy. Finally, comparison and ablation experiments on 25 instances demonstrate the superior performance of the EDA-TS algorithm in solving the SM-DFJSP, highlighting the effectiveness of the multi-population strategy and non-dominated solutions memory mechanism.